Saturday, June 11, 2011

Would you like 2 yaks or 3 yaks with that test?

I took a (small) risk last week and tried something different for our Year 8 Algebra test. I'm calling it the "2-Yak/3-Yak test". The idea in a nutshell: students choose the level of difficulty of the test.

Each section of the test provides questions grouped into level of difficulty indicated by the number of yaks. Students were required to do the 2-Yak column, and then for each section choose between the 1-Yak or the 3-Yak column. If students selected the 2-Yak/3-Yak combo, they would automatically get the marks for the 1-Yak questions. I suggested to students that if they wanted to do the 3-Yak questions but thought they might be too hard, to just do the 1-Yak/2-Yak, move on to the next section - and then at the end, if they had more time, go back and try some 3-Yak questions.

So why do this? And what did the students think?

Why did I try this approach?
  • I wanted to provide many harder questions to challenge the stronger students, without penalising other students with the negative feedback and discouragement that comes from getting poor test marks,
  • I didn't want the stronger students to waste time doing easy questions when they could be doing harder questions,
  • I wanted to see what would happen. My colleagues gave me puzzled "what-if...." looks when I showed them, and to be honest, I didn't quite know the answers to all their questions - and I wasn't 100% sure how the marking would work out. So I really was curious.
So what happened?
  • It really did work! The stronger students did indeed select the harder questions, and the less strong students tried some of them.
  • It turned out to be quite easy to allocate automatic 1-Yak marks to students who didn't do that column. In most cases I gave them full marks - although if I saw the student didn't really understand the concept I docked a mark. But it was very clear there was no need for these students to do the 1-Yak questions
  • I ended up giving all students two scores: a Baseline score (1-Yak+2-Yak or 2x 2-Yak) and an Extension Score based on the number of correct 3-Yak questions. In my markbook, I recorded these in two columns - treating them as two different tests. For my final grading, I will weight them so students doing the 3-Yak question get the equivalent of 0.5 marks extra per 3-Yak question.

What did the students think?
A quick anonymous student poll of the class showed overwhelming support, with 80% expressing strong support, and the remaining 20% saying "it was alright". No-one thought it was a bad idea. Here's a selection of student comments:
"Great - it was easier to choose what to do - I didn't panic as much"
"I found it very good because you knew what level you were doing"
"A great idea because it has given us the option to choose the level of difficulty and I really appreciate the method."
"It was great so people could be challenged a bit or do their personal level of work"
"I think that the last few questions in the 3 yak column should be harder (slightly)"
"I really like it. I think it's very helpful and the little pictures made me not as stressed."
"Because we can choose how hard or easy we want the test to be, according to our level of maths."
And here's my favorite:

Why yaks? We have been using a Himalayan mountain climbing metaphor for our algebra journey. Wonderful beasts and the students were fascinated by them.

Test Algebra v2

Test was created using Word 2010 - essentially just one one large multi-page table with cells merged, use of the new Office Equation editor (it's really come a long way these last few years) and some yak pictures.


  1. Love the idea... though I'd be worried about what the students think re. the 'final grading', where extension marks end up as normal marks. Transparency?

  2. Absolutely. I confess I was still working it out as I did it. Having done it the first time, I would tell the students in advance that doing the 3-Yak questions are worth 1.5 marks compared to 1 mark for the 1-Yak and 2-Yak questions. Maybe I should just be explicit in the marking at the top: show the Baseline Score, the Extension Score and the total calculated Grade mark (Baseline + 0.5 x Extension).

    However I should add this test is part of larger context of Standards Based Grading - where most of the class grade comes from repeatable quizzes, with what is close to a binary grade for each outcome: not demonstrated (C); on the way (B); mastered - with small errors (B+); mastered without error and extended (A).